Table 2 Validation protocol of the Healthy Food Intake Index (HFII) among pregnant Finnish women at high risk of GDM: type of validity and HFII components, and adopted approach for evaluation

Reproducibility

 1. Does the HFII₁ measured at 1st trimester adequately agree with the HFII₂ measured at 2nd trimester?

1. Kappa coefficients between 1st and 2nd pregnancy trimesters’ HFII components, intra-class correlation coefficient for 1st and 2nd trimesters’ total HFII.

Content validity

 2. Do the index components cover all the food groups of the underlying recommendations of healthy diet (NNR)?

2. Comparing the content of the HFII with NNR.

Construct validity

 3. Does the HFII create variation in the population?

 4. Is scoring independent from energy intake?

 5. Does the HFII have multidimensional construct and what are the dimensions it measures?

3. Item analysis of the HFII components: corrected item correlation and item mean.

4. Energy intake from food records. Comparisons between the HFII categories: general linear models

5. Iterated principal factor analysis for the HFII components matrix.

Components

 6. Do the components have independent roles within the HFII?

 7. Which components provide the highest and lowest scores?

6. Corrected item correlation

7. Item mean

Criterion validity

 8. Does a linear trend in nutrient intake exist across the HFII categories, or index component categories?

 9. Does the HFII distinguish between groups with known differences in diet quality?^a

  a. Age

  b. Education

  c. BMI

  d. Smoking

  e. Physical activity

8. Statistical comparisons between the HFII categories: bootstrap-type general linear models with the appropriate contrast

9. Same as 8^.